ABSTRACT Accurate breast cancer risk assessment has the potential to distinguish women at higher risk who need enhanced screening, preventive or risk-reducing therapies and surgeries from women at lower risk who can be spared interventions that yield little benefit and may cause harm. Breast cancer risk assessment is currently hampered by limited precision and accuracy of existing risk prediction models. Women with a family history of breast (FHBC) have the greatest need for better risk assessment as they often receive the same clinical recommendations despite substantial heterogeneity in the underlying risk by the extent of FHBC. Integration of mammographic breast density (MBD), a strong and readily assessable risk factor for breast cancer, in combination with detailed FHBC data, offers an exceptional opportunity for enhancing existing risk assessment methods. MBD generally declines with age, but the rate of change varies considerably between women, and may be particularly important to breast cancer risk; we have demonstrated that women who remain at high MBD over time are more likely to be diagnosed with breast cancer than women whose MBD decreases. Attempts to improve risk prediction models by incorporating MBD has had limited success as studies have used one-time measures of MBD and included mostly postmenopausal women for whom the largest changes in MBD may have already occurred. We propose to investigate within-individual changes in MBD over a 10- year period in relation to incident breast cancer by the extent of FHBC (Aim 1), and evaluate how changes in MBD may improve several clinical risk prediction models (Aim 2) and clinical risk stratification forming the basis for risk-based surveillance and preventive care (Aim 3). We will address these aims by building upon the U.S. Sister Study, a prospective cohort of 50,884 women with one or more sisters diagnosed with breast cancer who were personally breast cancer-free at enrollment in 2003-2009; active annual follow-up is conducted for at least 10 years with each woman. Using a nested case-control design, we will retrieve existing mammograms for all incident breast cancer cases diagnosed at ages ? 60 years (n=1,242 cases to date) and controls matched on age and enrollment year (2 controls selected per case at the time of case identification). We will undertake a comprehensive assessment of MBD, using both clinically available qualitative measures used in clinical practice, and assessing quantitative measures that allow for measurement of smaller changes and different components of MBD (e.g., dense, nondense tissue) that are independently associated with breast cancer risk. Our team?s experience in leading studies of MBD and prospective available data in the Sister Study will afford an unparalleled investigation of prospective MBD changes in relation to breast cancer risk, modifiable and genetic factors, and how to use this information to enhance risk assessment in women with FHBC. These results are necessary to inform personalized risk-based surveillance and prevention programs.